Narrative:

On jan, 1994, at XA30Z the aircraft I was 'captaining' passed through our assigned altitude of FL260 and leveled at FL250. At the time of the incident, we were 85 mi east of roc en route from kbaf to kbuf on the first of 3 legs. The aircraft was on the autoplt with navigation and flight level change selected for the descent from FL410 into the approach. The WX in buffalo was 3 scattered 5 overcast with visibility down to 1/2 mi in light rain and fog. Prior to the descent clearance to FL260, we had picked up the ATIS anc completed the descent checklist. I was reviewing the approach procedure in preparation for the crew briefing when the clearance to descend was issued. Upon receipt of the clearance to FL260, first officer read back the clearance and reset the altitude selector. I looked up to confirm that he had set FL260, which he did, and then reset the autoplt to recommence the descent. I then proceeded to give the approach briefing to the first officer, during the descent, which he acknowledged. As we leveled, the first officer was setting up the ILS qdm and dialing in the appropriate frequency. I was adding power to maintain speed in the level off when ATC cleared us to continue the descent and call altitude leaving. It was then that we realized we had leveled at FL250 which was now set in the window. We are still unsure of when the change from FL260 to FL250 occurred or how it occurred. This incident has identified 2 problems that we have to correct or compensate for as a flight crew. Firstly is the tendency in today's automated flight decks to be busy with tasks other than the primary one of flying the aircraft, especially during climb and descent. Sterile cockpits below 10000 ft is not good enough anymore. The more capable the aircraft, the more vigilance seems to be required. The other problem is a design related shortcoming in the actual altitude selector function of SPZ8000. Unlike older selectors with ratcheting or mechanical resistance, this selector is extremely smooth. The result is that is can be difficult to set and is easily bumped off its setting when the course or heading knobs are moved. This is especially true when encountering turbulence or during heavy cockpit workloads. To fix the problems, we have instituted new cockpit resource management procedures. Firstly, we will not allow for nonessential tasks to be performed during the transition to or from cruise flight and at no time will both pilots have their attention drawn away from the flight instrument. Secondly, the pilot setting the altitude in the selector will announce that it is set and then announce that it is indicated after both pilots confirm its correctness. Lastly, every effort will be made to complete approach briefings and setups during level flight where the PNF can concentrate fully on that task and the PF can monitor same. Supplemental information from acn 262036: callback conversation with reporter revealed the following information: the knob is also positioned in between the course select and heading select knobs on the first officer's side. Therefore, the knob could easily be bumped to a different altitude while turning either of the other 2 knobs.

Related reports:

Google
 

Original NASA ASRS Text

Title: ALT BUST.

Narrative: ON JAN, 1994, AT XA30Z THE ACFT I WAS 'CAPTAINING' PASSED THROUGH OUR ASSIGNED ALT OF FL260 AND LEVELED AT FL250. AT THE TIME OF THE INCIDENT, WE WERE 85 MI E OF ROC ENRTE FROM KBAF TO KBUF ON THE FIRST OF 3 LEGS. THE ACFT WAS ON THE AUTOPLT WITH NAV AND FLT LEVEL CHANGE SELECTED FOR THE DSCNT FROM FL410 INTO THE APCH. THE WX IN BUFFALO WAS 3 SCATTERED 5 OVCST WITH VISIBILITY DOWN TO 1/2 MI IN LIGHT RAIN AND FOG. PRIOR TO THE DSCNT CLRNC TO FL260, WE HAD PICKED UP THE ATIS ANC COMPLETED THE DSCNT CHKLIST. I WAS REVIEWING THE APCH PROC IN PREPARATION FOR THE CREW BRIEFING WHEN THE CLRNC TO DSND WAS ISSUED. UPON RECEIPT OF THE CLRNC TO FL260, FO READ BACK THE CLRNC AND RESET THE ALT SELECTOR. I LOOKED UP TO CONFIRM THAT HE HAD SET FL260, WHICH HE DID, AND THEN RESET THE AUTOPLT TO RECOMMENCE THE DSCNT. I THEN PROCEEDED TO GIVE THE APCH BRIEFING TO THE FO, DURING THE DSCNT, WHICH HE ACKNOWLEDGED. AS WE LEVELED, THE FO WAS SETTING UP THE ILS QDM AND DIALING IN THE APPROPRIATE FREQ. I WAS ADDING PWR TO MAINTAIN SPD IN THE LEVEL OFF WHEN ATC CLRED US TO CONTINUE THE DSCNT AND CALL ALT LEAVING. IT WAS THEN THAT WE REALIZED WE HAD LEVELED AT FL250 WHICH WAS NOW SET IN THE WINDOW. WE ARE STILL UNSURE OF WHEN THE CHANGE FROM FL260 TO FL250 OCCURRED OR HOW IT OCCURRED. THIS INCIDENT HAS IDENTIFIED 2 PROBS THAT WE HAVE TO CORRECT OR COMPENSATE FOR AS A FLC. FIRSTLY IS THE TENDENCY IN TODAY'S AUTOMATED FLT DECKS TO BE BUSY WITH TASKS OTHER THAN THE PRIMARY ONE OF FLYING THE ACFT, ESPECIALLY DURING CLB AND DSCNT. STERILE COCKPITS BELOW 10000 FT IS NOT GOOD ENOUGH ANYMORE. THE MORE CAPABLE THE ACFT, THE MORE VIGILANCE SEEMS TO BE REQUIRED. THE OTHER PROB IS A DESIGN RELATED SHORTCOMING IN THE ACTUAL ALT SELECTOR FUNCTION OF SPZ8000. UNLIKE OLDER SELECTORS WITH RATCHETING OR MECHANICAL RESISTANCE, THIS SELECTOR IS EXTREMELY SMOOTH. THE RESULT IS THAT IS CAN BE DIFFICULT TO SET AND IS EASILY BUMPED OFF ITS SETTING WHEN THE COURSE OR HDG KNOBS ARE MOVED. THIS IS ESPECIALLY TRUE WHEN ENCOUNTERING TURB OR DURING HEAVY COCKPIT WORKLOADS. TO FIX THE PROBLEMS, WE HAVE INSTITUTED NEW COCKPIT RESOURCE MGMNT PROCS. FIRSTLY, WE WILL NOT ALLOW FOR NONESSENTIAL TASKS TO BE PERFORMED DURING THE TRANSITION TO OR FROM CRUISE FLT AND AT NO TIME WILL BOTH PLTS HAVE THEIR ATTN DRAWN AWAY FROM THE FLT INST. SECONDLY, THE PLT SETTING THE ALT IN THE SELECTOR WILL ANNOUNCE THAT IT IS SET AND THEN ANNOUNCE THAT IT IS INDICATED AFTER BOTH PLTS CONFIRM ITS CORRECTNESS. LASTLY, EVERY EFFORT WILL BE MADE TO COMPLETE APCH BRIEFINGS AND SETUPS DURING LEVEL FLT WHERE THE PNF CAN CONCENTRATE FULLY ON THAT TASK AND THE PF CAN MONITOR SAME. SUPPLEMENTAL INFO FROM ACN 262036: CALLBACK CONVERSATION WITH RPTR REVEALED THE FOLLOWING INFO: THE KNOB IS ALSO POSITIONED IN BTWN THE COURSE SELECT AND HDG SELECT KNOBS ON THE FO'S SIDE. THEREFORE, THE KNOB COULD EASILY BE BUMPED TO A DIFFERENT ALT WHILE TURNING EITHER OF THE OTHER 2 KNOBS.

Data retrieved from NASA's ASRS site as of July 2007 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.